INDEX

Explanations

turned to

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ious

-0.24

åĪĹè½¦

-0.23

æľīä¸ĢçĤ¹

-0.23

è¾¹

-0.23

ãģĹãģı

-0.23

imus

-0.22

ingga

-0.22

é¡ºä¾¿

-0.21

minutes

-0.21

å¤ĩæ³¨

-0.21

POSITIVE LOGITS

æ±ĤåĬ©

0.31

å¯»æ±Ĥ

0.30

ä¹ĭä½ľ

0.26

å¤§æµ·

0.24

 Ð¸ÑģÑĤÐ¾ÑĩÐ½Ð¸Ðº

0.22

 fellow

0.22

 recourse

0.22

å¦Ĥä½ķçľĭå¾ħ

0.22

éĢĤåĲĪèĩªå·±

0.22

 refuge

0.22

Activations Density 0.032%